SolidWorks first foray into the cloud has been n!Fuze, an online sharing and markup app. It is widely seen as a flop as a product. Problems with the implementation and the marketing have left a lot of people wondering how SolidWorks saw this as an improvement over 3D Instant Website, which was discontinued just months before n!Fuze’s release. Still, some folks are eager to know more about cloud pricing, and now that we have seen a real cloud app from SolidWorks, albeit a dysfunctional one, we have something concrete upon which to base our usually baseless conjecture. n!Fuze is $70/month, requires 2 licenses to run (one for creator one for viewer), but only priced in 2 different time chunks: 12 months for $840 ($70/mo) or 3 months for $231 ($77/mo).
Matt West (only the messenger here) is the sacrificial lamb SolidWorks sends to defend them here on my blog without any new information. Matt’s a decent guy, but he’s defending what looks to be a more and more ridiculous situation, as time goes on. SolidWorks doesn’t offer any new information about cloud plans, while customers range from the restive to the resigned.
it may turn out that there will not be an economically viable business atmosphere in which to offer full-blown CAD for heavy duty use over the internet for even the next decade.
Here’s the argument, or one argument. If you don’t need something very much, you can rent it. That’s a common thing. I rented a hydraulic post hole digger because I only needed it for an afternoon, I wasn’t going to use it to make a living. Matt’s on solid ground so far.
But when you consider that n!Fuze is only offered in chunks of 3 or 12 months, it requires at least two paying licenses to use it, and you extend the pricing to SolidWorks sort of functionality, it could really start to add up. Say that SolidWorks has 50 times the functionality of n!Fuze (it easily has much more), lets be generous and say that SW would only be $250/month instead of $70x2x50=$7000. But you can’t just rent it for a day or a week. You have to rent it in chunks of 3 or 12 months, and by that time, you may have spent a couple thousand dollars.
If you’re going to spend that kind of money, you might as well just hire some chap in India to do the work for you for $15/hr. So SolidWorks online will sell a lot of outsourcing. Why pay that much for software, training, and an operator when you can go straight to the finished CAD data for much less? Even at engineering consulting rates of $75-$100/hr, consulting with a real pro would still be an attractive alternative to renting hardware/software. With a consultant you’re going the extra distance and also renting the operator, but only getting it all for the exact amount of time that you need. I mean, consultants don’t charge you for 3 months work, but only show up for 1 week, that’s ridiculous. Why should you “rent” software for any longer than you need it, it just doesn’t make sense.
I mean, let’s be clear, hardware is cheap these days. You can get a very passable box to run SolidWorks for $1500. That investment will last you 3 years. The software, depending on what you need is at least $4000. That’s a one-time investment. The operator is needed in both cases, but in a short term rental situation, it may be difficult to hire an experienced operator.
If you rent, at $250/month, in a year you pay $3000. At the end of the year all you get is a door slamming on your back. You might have data files (and you might not, if V6 is any indication with the database), but you won’t have access to the software any more. If you purchased, you would have your software nearly paid for at that point. After 2 years of renting, you have spent $500 more than purchase, and after 3 years of renting, you have spent $3500 (per seat) with absolutely nothing to show for it, and no “rent to own” value or buy out option. This doesn’t account for the extra cost of a heavy duty, extra reliable internet connection.
So, if the pricing scheme stays the way it has indicated with this first foray, I don’t think SolidWorks has a chance to sell much cloud. It would be a full-on cloudy car-wreck on the information highway.
There’s a reason why we aren’t getting any information about SolidWorks cloud plans: I think it could be because they still don’t know themselves. This whole thing is complicated, and beyond the new programming and the bugs it will bring, beyond the technology, beyond the argument of whether it is secure enough, beyond the argument about who owns the data and how to handle risk or breach, it may turn out that there will not be an economically viable business atmosphere in which to offer full-blown CAD for heavy duty use over the internet for even the next decade. If SolidWorks really doesn’t have this figured out nearly 2 years after the initial public announcement, I think the whole idea will at least be scaled back severely in order to do anything at all with it.
There’s another scheme, but it doesn’t involve extorting a lot of money from customers for a service that may not be a very good idea to begin with. It might actually allow customers to only buy what they need (gasp!!) There would be no maintenance, no guaranteed annual income, no need to buy services you don’t need and don’t use from resellers who can’t/won’t/don’t deliver. So I’m quite sure SolidWorks will not go this way. Autodesk also does not appear to be headed this way, they are plowing forward with a cloud subscription model. But let’s take a look at it anyway, and see what a company that respects its customers might do.
Take a look at this site, just as an example. I don’t have any attachment to these people, I just found them in a web search for “render farm service”. Rendering is something we have said several times would be a great application for distributed computing, and distributed computing is a form of cloud. Anyway, the pricing scheme here is not in huge chunks of time, it’s in Ghz hours. That’s processor clock speed x number of processors (cores) x number of hours to complete a job. So you just pay for what you use. Fair and simple.
(By the way, if you really do need a render farm’s services and you use Modo for rendering, Anna Wood has a service called RenderBay where she charges reasonable rates, on a pay for what you need basis.)
With a scheme like that, Matt’s assertion that you could just rent out what you need makes a lot more sense. But honestly, I don’t see SolidWorks doing pricing on a real time basis. They’ve made a business of selling people much more than they need for a long time, and I don’t see that changing. Just like the cell phone company selling you X minutes and you only ever use .6xX.
What I’m finding is that in reality, people range all over the map on the whole CAD in the cloud issue. There are a lot of “wait and see” folks out there. There are also a lot of people ready to drop maintenance at the next opportunity until this sorts itself out. The few folks who are upbeat about the cloud tend to be hawkish on the business side, (and they all have something to sell) looking to drive a hard bargain with people who they assume to be so desperate to touch SolidWorks that they will ignore any financial common sense that ever existed. There are those who simply believe “It’s the future!” (like $5 ATM access fees and 39% interest on credit cards – just roll over and take it). My position seems to change over time. Right now I’m of the opinion that it will be a decade before we see full-blown 3D CAD on the web. In the mean time, CAD In The Cloud will be scaled back to look like everything else that’s on the cloud – smaller, simpler, supporting applications without the huge underlying data and super intense graphics. CAD data on the cloud? You know, there might be some people willing to accept that, but I think the numbers will be small.
We’ve said a number of times the things that make sense for cloud or distributed computing in general: rendering, FEA, CFD, and stuff like that. Intense compute, distributable, controllable. I think SolidWorks is going to have to scale back and focus on these first. Of course they’re going to need to fix n!Fuze even before that. They could start by renaming it, and cutting an order of magnitude off the price, and make the rental period more granular.
I firmly believe at this point that Dassault has a big misconception about what the typical SolidWorks user is all about. When I worked with SolidWorks resellers, more than half of our sales came from small installations, 1-3 seats, and I doubt that ratio has changed much. All of this cloud stuff only makes sense for big installations, and then not Amazon cloud, but local cloud. I’ve talked to a lot of people within SolidWorks, from the resellers, from prominent users, CAD journalists, and self-appointed wonks. The most optimistic of them are cautious. The center seems to be wary of the changes.
I’ve heard the conjecture that SolidWorks is really being pushed to the cloud because there will be no way for semi-computer-literate end users to maintain their own installations. The V6 install will be based on Enovia and its databases. This is the type of install that you have professional IT people or a consultant to do. No wonder SolidWorks is now arguing that moving to the cloud eliminates the need for all that IT overhead – the current installation doesn’t require much, but the V6 one is sure to. It’s a problem that causes itself. This is the wrong direction for the typically independent SolidWorks user. Most of us bought SolidWorks because it simplified things,and V6 will severely complicate things, so they have to take it out of the end user’s hands and put it on the cloud. This is less based on conjecture than you might think. I’ve heard all of this in one form or another from more than one source.
To me, the real question is “when will SolidWorks/Dassault realize the problems inherent in this direction? I think they already know at least some of it. They are out on a limb already. The question is are they going to swallow some pride and come back to safety or are they going to start sawing?